I don’t believe there’s a real
difference between the integer version and the short text version, but I believe
the short text will be faster than the long text. I don’t have any
performance numbers to prove how much or if I’m even right, sorry.
Matt
From: [email protected]
[mailto:[EMAIL PROTECTED] On Behalf
Of jgserrault
Sent: Thursday, May 05, 2005 8:36
PM
To: [email protected]
Subject: [flexcoders] Performance
question relating to array element referencing
I am about to embark on building a component that will involve heavy
usage of arrays, and before I write a benchmark
routine to answer my
question, I thought I would see if anyone already
had the answer. My
question is basic: When referencing array
elements using the
following approaches/methods which is the fastest
to execute, and
then what is the next best method relative to
best? For example:
Method x is best, method y would be about 1.4x,
and method z would be
1.8x.
Method A (Integer ordinals from 0 or 1 to Length):
Array[0]
Array[1]
Array[2]
Method B (Text ordinals short in length):
Array[i00]
Array[i01]
Array[i02]
Method C (Text ordinals long in length):
Array[ArrayItemDescriptionThatIsVeryInformative1]
Array[ArrayItemDescriptionThatIsVeryInformative2]
Array[ArrayItemDescriptionThatIsVeryInformative3]
I am just looking for a realistic approximation so
I can make an
informed decision. Obviously Method C offers
the greatest
flexibility, but if the execution price is 2+
times the quickest
method then it needs to be reconsidered...at least
in my intended
usage.
Also, in the case of text based ordinals are there
ranges that are
better to use than others (i.e. a-z or A-Z).
Sorry if this has been
covered in a tech doc.
Yahoo! Groups Links
|
